It is well known that matrix displays such as liquid crystal displays, both passive matrix and active matrix varieties, are composed of two planes (usually clear glass or plastic) having a multitude of conductive electrodes which sandwich a film of electro-optic material, such as liquid crystal material. Each point of intersection of the conductive electrodes between the front and back planes forms the site of a picture element (pixel). In the active matrix display varieties, a thin film of non-linear or active devices such as diodes, transistors, or varistors are also included at the intersections of the electrodes.
It is understood that liquid crystal displays (LCDs) are activated by an AC wave form in order to minimize destructive effects to the display element which are caused by accumulating DC bias. These destructive effects consist of electrolytic plating and chemical breakdown of the electrodes and of electrochemical breakdown of the crystal material.
Thus, prior art liquid crystal display elements in the "on" state are alternately subjected to equal and opposite polarities of electrical bias on a continuous basis at a fixed frequency (AC drive) to avoid the destructive properties of an accumulated DC bias. Several compromises are made to drive LCDs with this existing scheme. Higher AC drive frequencies allow the display to respond more quickly to an update, but have lower display contrasts, narrower viewing angles and use power less efficiently. Lower AC drive frequencies are more efficient and have greater contrast, but update more slowly since the AC drive frame cycle must always be completed before an update is made in order to achieve a neutral bias.
The present invention avoids these compromises by means of an improved driving scheme which eliminates the burden of requiring frequent and symmetric reversals of drive polarity. This enables the implementation of improved DC drive techniques while still neutralizing the DC bias on the pixels before destructive effects occur. The present invention allows the display controller to respond more quickly to display update requests by eliminating the need to complete the current frame cycle and the opposing polarity cycle before responding to the next display update request. When the display controller must change the gray level of one or more pixels, this request is acted upon immediately, with the existing DC bias of the display element stored in memory so that the display element's bias can be compensated for at a later time. This technique is called "bias reconciliation". The net DC bias on the display element at the time of update is called "DC bias violation". With the present invention, when the display controller receives a request to update the display, the controller does not need to delay until the current display frame cycle is completed, as is required by prior art display drive systems. Rather, the circuit reacts instantly to the update, and the DC bias violation status of the display element is updated in memory.
"Real Time Display Simulation" refers to the use of memory and computation means to simulate the condition of the display in real time. Aspects of the display which are simulated in the present invention include the existing electro-optical condition of the pixels, the accumulated DC bias on the pixels, and the difference between the existing condition of the display and the most recent demanded image. Use of real time display simulation techniques allows the implementation of the display drive and control techniques which will be explained herein.
"DC bias violation" is a represented quantity referring to the integration of the varying voltage levels applied over time to each individual pixel. "Bias reconciliation" is the reduction and neutralization of the DC bias violation to insure the maintenance of safe DC bias violations. This is achieved by means of keeping track in memory of the accumulated electrical bias on the pixels and reversing the polarity of the drive signal before any pixel reaches a predetermined bias level. The bias status can, therefore, remain or accumulate in one polarity for multiple display periods. (For prior art, the polarity of the drive signal is reversed at a fixed frequency that is between every 400th to 30th of a second).
"Maximum bias violation tolerance" (MBVT) refers to a transfer function of time and DC bias. It is the measure of the net DC bias a pixel can sustain without suffering irreversible damage due to electrochemical reactions. (Note: With existing fixed cycle AC multiplex drive methods, the pixels experience non-zero DC bias within a fixed frame cycle, but this is always brought to zero by the end of the frame cycle.) MBVT refers to the upper limit of the DC bias violation a pixel can sustain. Exceeding the MBVT for a display element will cause destructive effects to the display and will lower the life expectancy of the display. The parameters for MBVT will vary among different displays as a function of the materials used and the structure of the display.
"Selective Real Time Drive Sequencing" is the display control technique in the present invention in which the display controller selectively varies in real time the electrode drive sequence, the duty cycle, and the backplane/segment plane drive functions. "Pixel Power Modulation" is a novel display control technique in the present invention in which the display controller selectively varies (or modulates) in real time the power applied to individual pixels to maintain them in the desired gray band. The employment of these techniques enables improved and more flexible means of driving and controlling passive and active matrix liquid crystal displays.
A passive matrix liquid crystal display can be viewed as a matrix of slightly leaky capacitors as illustrated in FIG. 2. Each matrix location is identified by a corresponding equivalent resistance-capacitance pair Rnm Cnm where n is the row location and m is the column location. There is minimal resistance (approximately 100 ohms) in the connections between pixels, so when a charge is established across a pixel, it dissipates quickly through the matrix.
An active matrix liquid crystal display can be viewed as an array of capacitors all having the backplane as a common plate with each X and Y column and row location having an individual active element in contact with the active plane, as illustrated in FIG. 3. FIG. 3 shows both front and side views of an active matrix LCD display. Therefore, the pixels in both passive and active matrix LCD's require repetitive rewrites in order to maintain them at desired gray levels or to drive them to new gray levels. To accomplish this, prior art multiplex drives operate as follows: The drive signals are applied to two sets of electrodes typically arrayed in rows and columns. Voltage select signals are sequentially and periodically applied to each backplane electrode in a repetitive cycle. In synchronism with the backplane electrode select signal, the segment plane select signals are applied in parallel, thus affecting the electro-optic conditions of the pixels at the intersections of these selected backplane electrodes and the selected segment plane electrodes. An ON pixel has, therefore, experienced an applied RMS voltage that exceeds its threshold turn-on voltage, whereas an OFF pixel has experienced a voltage below threshold voltage.
The frame refresh rate must be kept above 30 Hz or the display will appear to flicker. As the number of display elements increases in the prior art, the multiplex ratio must be increased in order to address the greater number of elements. As the multiplex ratio is increased (more backplanes), less time is available to sequentially drive each backplane, and the driver must operate at a higher voltage and frequency to produce the required RMS drive signal.
As the drive voltage is increased and time duration decreased in response to these higher multiplex ratios, the discrimination ratio (the difference) between on and off RMS voltage decreases. This creates an appearance of semi-selected pixels, decreases display contrasts and introduces cross talk. In effect, since there is less time available to drive each pixel, there is less controllability, thereby decreasing the number of available gray levels.
Another major intrinsic drawback of the prior art fixed cycle AC matrix drive techniques is illustrated as follows:
At the point in the driving cycle at which the pixel is driven to the opposite polarity it passes through the zero voltage condition. This causes the opacity of the pixel to decrease (become less gray) until it reaches the full off condition for a brief moment. The pixel then becomes more gray as it is driven to the reverse polarity. The gray level perceived by the eye is thus less intense since the eye integrates all of the gray levels of the transition. This point in the drive cycle exhibits a decrease in display output in conjunction with an increase in energy consumption. This effect is counter to what is desired and represents a major difficulty. This decreased display output effect becomes progressively worse as the drive frequency is increased.
The purpose of a display drive controller is to cause an image to appear on the display which conforms as closely as possible to a demanded or desired image. In many displays (e.g. CRTs, LCDs, LEDs, etc.) the projected image is not static like a photograph. Such displays are referred to as monostable. In monostable displays, even when the desired (or demanded) image is unchanging, the gray level of each pixel is continually varying in intensity. Typically, the gray level of each pixel decays until it is refreshed or driven, which "recharges" the pixel to a higher gray level.
In LCD drive controls, the drive's refresh and decay mechanism is employed to achieve desired gray levels as follows. Each pixel is white in appearance at the low energy state (also called the ground state or off state). Each pixel appears black at the high energy state (also called the saturation state or on state). At energy levels between the low energy white state and the high energy black state, liquid crystal pixels will display a range of gray levels. However, prior art LCD controllers do not take advantage of that range of gray levels as will be described hereinafter. (Note: By altering the orientation of the polarizing filters used in a liquid crystal display, the display's appearance can be reversed so that a pixel appears black in the low energy state and white in the high energy state. For the sake of clarity, this discussion will proceed with the assumption that the pixels appear white in the low energy state. However, the present invention can be applied to displays with either orientation.)
To increase the gray level of a liquid crystal pixel, the drive controller applies an electric field across the pixel. The field distorts the molecular orientation of the liquid crystal material thereby changing its optical characteristics, appearing as an increased gray level.
A relationship exists between the voltage across a liquid crystal pixel and the gray level of the pixel. The curve which shows the relationship between applied voltage and gray level is called the "electro-optic turn on curve". Similarly the pixel exhibits an "electro-optic turn off curve" as the liquid crystal material relaxes and returns to its low energy state when the applied voltage is removed. It is noted in U.S. Pat. No. 4,921,334, Matrix Liquid Crystal Display With Extended Gray Scale, by Boris A. Akodes, that the distribution of these gray levels is not linear. Additionally, the "electro-optic turn on curve" is not symmetric to the "electro-optic turn off curve"--the hysteresis of the turn off curve is typically 21/2 to 4 times slower than the turn on curve time characteristic. (FIG. 4 of this application illustrates the nonlinearity of the turn on curves.) The time required for liquid crystal material to undergo molecular twist from an off state to an on state is referred to as the "excursion time".
Typical full on excursion times for LCD displays with current materials range from 0.05 milliseconds for ferroelectric material to 60 milliseconds for supertwisted nematic material at room temperature, depending on the particular liquid crystal material used. This is the time delay required for an element to change from a fully off state (white) to a fully on state (black) when driven by an RMS voltage exceeding its threshold turn-on voltage.
Several factors affect the voltage/gray level transition curve characteristics. These factors are inherent in the design and construction of the display and in the ambient environment of the display. Among the inherent factors affecting the voltage/gray level transition curve are:
1. Material Characteristics PA1 2. Display Design PA1 3. Ambient Conditions PA1 Characteristic 1. Any net DC bias on the pixels must be neutralized within one drive cycle or one frame cycle; PA1 Characteristic 2. Only one backplane can be selected at a time; PA1 Characteristic 3. The backplanes must be driven sequentially in a regularly repeating frame cycle; PA1 Characteristic 4. The functions of backplane and segment plane in the rows and columns are fixed; that is, they can not be interchanged selectively in real time; PA1 Characteristic 5. Gray levels are produced by generating set proportions of full on signals (i.e. at or above the saturation voltage) and full off signals (i.e. below the threshold voltage) at a given pixel on a frame by frame basis ("interframe modulation") PA1 1. Use of memory means to store and update information on the display elements such as accumulated bias and present gray level; PA1 2. Use of real time control and simulation techniques to calculate optimal or near-optimal drive signal patterns in real time; PA1 3. Use of power modulation techniques which allow a greater variety of voltage levels to be used to generate the additive and subtractive drive signal levels to be used to drive the pixels; PA1 4. Selective drive signal means enabling non-sequential and multiple addressing of the electrodes and interchange of the functions of backplane and segment plane between the row and column electrodes.
a. Electro-optical characteristics of the particular liquid crystal material PA2 b. Electrical characteristics of the barrier layers between the electrodes and the liquid crystal material PA2 c. Electrical resistance of the electrodes PA2 d. Viscosity of the liquid crystal material PA2 e. Elasticity constraints. PA2 a. Thickness of the liquid crystal film in the display spacers in the display PA2 b. Size, type and placement of the PA2 c. Alignment angle and anchoring characteristics of the liquid crystal film and the barrier surfaces PA2 d. Area and layout of the individual pixels PA2 a. Voltage of the applied drive signal PA2 b. Existing gray level of the pixel PA2 c. Ambient temperature of the display
It is important for the designer of a display driver to understand these inherent characteristics of a display that influence the shape of the electro-optic turn on and turn off curves. Such an understanding helps to design a driver which offers improved display quality and image predictability.
All existing LCD controllers, including the present invention, are open loop controllers (i.e. the display controller has no feedback from the actual display). One of the innovations of the present invention is the simulation in real time of the characteristics outlined above. The display controller refers to the real time simulation to obtain key information to determine the drive signals for the display. This allows the impact of these characteristics to be included in the computations used to determine the drive signals for the display. Proper use of the simulation allows a greater number of pixels to be driven to a greater number of gray levels with greater accuracy. Employment of the present invention in color displays will allow a greater number of colors to be displayed with greater accuracy.
Throughout this patent application, reference is made to "real time" driving and/or control signal generation. "Real time" means that the drive signals are applied as generated by the control system as a continuous response to the most recent demanded image. Additionally, the present invention can apply drive signals to the array of rows and column electrodes "asynchronously", which means that there is not a preset sequence of activating rows or columns. The requirements of timing cycles, frame sets, or preset sequencing cycles as practiced in prior art for controlling the application of the control drive signals to the pixels, and for assuring that all DC bias is neutralized within one frame set, are not necessary in the practice of the present invention. This is made possible in the present invention by the use of real time computations and memory storage means which enables display simulation and DC bias tracking.
In sum, the prior art generally drives row and column electrodes in a predetermined sequence and according to a clock synchronized with the prior art AC signal. The present invention allows pixels to be driven selectively and in any sequence (e.g. synchronous, asynchronous, multiple backplanes selected, skipped backplanes), The order in which the pixels are driven is determined by underlying principles of this invention.
Prior art display drive controls, when faced with the requirement of ever increasing numbers of pixels to control, have adopted an approach of driving the display harder. That is, higher voltages are used with faster drive signals, as described above. This approach to servicing increasing numbers of pixels is a result of considering the LCD as an RMS responding device driven by AC wave forms. As will be described hereinafter, the present invention operates LCDs as DC voltage integrating devices. This approach overcomes several limitations inherent in the RMS responding approach. As discussed previously there is a limitation on the number of pixels which can be controlled and the number of gray levels which can be displayed using these prior art control schemes. These limitations are explained in "Scanning Limitations of Liquid-Crystal Displays", by Paul M. Alt and Peter Pleshko, IEEE Transactions on Electron Devices, February, 1974, pages 146-155, and "Reduction of Brightness Non-Uniformity in RMS Responding Matrix Displays", by T. N. Ruckmongathan, P. H. Verheggen, and Th. L. Welzen, Proceedings The Society for Information Display, Sep. 25-27, 1990.
As described in these papers, the number of backplanes in a display increases as the number of pixels increases. Servicing an increasing number of backplanes using prior art dictates a decreased amount of time available to service each backplane, and the decreased time available results in a corresponding decrease in controllability.
The present invention is not constrained by this trade off between number of backplanes and controllability. As will be shown, the amount of time available to service each backplane does not necessarily have to decrease as the number of backplanes increases.
The present invention eliminates some operating characteristics of prior art display controllers. These are:
The means by which the present invention eliminates the above characteristics, and the associated implications for improved display quality are described hereinafter.
An object of this invention is to provide a display drive control which operates without being hindered by any of the above prior art operating characteristics.
Another object of the invention is to provide an LCD display with improved imaging capabilities.
Still another object of this invention is to provide such an LCD display in which contrast, viewing angle and imaging capabilities as well as animation capabilities are improved over the prior art.
Another object of the present invention is to provide such an LCD display which reduces power consumption in relationship to the achieved image, is more versatile, provides greater clarity, increases the life of the LCD elements and is able to handle a larger number of pixels.
Still another object of this invention is to drive the pixels as DC voltage integrating devices.
Other objects and advantages and features of this invention will become more apparent hereinafter.